There's been a Yandex leaks lately like Have you checked any of that the buzz That's going around right now is on a Thing called uh bm25 and it's very Similar to TF IDF it's basically TF IDF Plus word count looking at the word Count of the other things that you're Comparing against and it's all over the Factors that were released it's kind of Interesting because you have to think That Yandex was built as a Google clone And I'm sure they got several Engineers That worked at Google and I saw a Statin I don't know if this is true or not the Results are about 70 the same you can See that maybe they're a few years Behind or they're kind of slight but if That's pretty close that's not too far Away and then so they're going in on This bm-25 as how they're doing it and What it is it's a bag of words a way to Do it so it doesn't care about grammar It doesn't care about how close words Are together it's simply taking all the Words and it's looking at term frequency Which is really a way that we attack Google now anyway